Best-Performing Value Strategies - Part 4: Backtesting The Magic Formula

by: Collin Moshman


Greenblatt's Magic Formula screens for inexpensive, high-quality companies.

The Magic Formula beats the market, including when restricted only to large caps.

We can improve on magic formula performance by ignoring the quality component and only buying cheap companies.

In the first three parts of this series, we looked at the results of mechanically buying low P/E companies; high F Score companies; and companies trading at less than liquidation value.

Today we look at one of the best-known value strategies: Joel Greenblatt's Magic Formula from The Little Book that Beats the Market.

The goal behind the Magic Formula is to identify high-quality companies trading at a low price. There are some different interpretations of the specifics, but the general process is as follows:

1. Take all companies in a particular universe excluding financial companies, utilities, and ADRs.

2. Rank the cheapest company 1, the second cheapest company 2, etc.

3. Rank the highest-quality company 1, the second highest-quality company 2, etc.

4. Select the N companies that have the lowest sum of the two ranks.

Defining "Cheapest" and "Highest Quality"

The Magic Formula measure of cheapness is:

Earnings Yield = EBIT / Enterprise Value

(EBIT = Earnings before interest and taxes)

This is a good choice. As we saw in the article on P/E ratios and alternatives, EV/EBIT has been the highest-performing price ratio since 2000.

The measure of quality is:

Return on Capital ((NYSE:ROC)) = EBIT / (Net Fixed Assets + Net Working Capital)

The denominator of ROC is net PP&E (i.e. net of accumulated depreciation) + current assets - current liabilities.

The higher a company's ROC, the more money it generates for each dollar of capital used to run the business, indicating a high-quality company.

Performance of the Magic Formula

The Magic Formula sounds good in theory, but does it work?

The following is the performance of a portfolio of the 100 best-ranked Magic Formula stocks from the Russell 3000 since Jan 1, 2000, with annual rebalance.

Magic Formula Backtest Results

This strategy has outperformed the benchmark with annualized returns of approximately 13% since 2000. The Magic Formula appears to do a good job of identifying high-quality companies at a cheap price.

Improving on the Magic Formula

The first potential way to improve on Magic Formula results is to change the measure of cheapness and/or quality. I was not able to improve on Magic Formula results by changing either. As an example, two methods I tried were switching earnings yield to the inverse of the P/E ratio, or switching return on capital to return on equity. Both lowered performance.

What if instead we rank only on one of the two magic formula criteria? If we rank companies purely on their return on capital, this strategy decreases annualized returns to 9.20%.

If instead, however, we only rank companies based on their earnings yield, we get an increase in annualized return to 16.47%.

When applied to Russell 3000 companies this century, at least, the Magic Formula is essentially taking two market-beating strategies and combining them, dragging down the performance of the superior one (cheap companies) with the performance of the inferior one (quality companies).

This is not an original observation on Magic Formula investing. See, for example, the discussion of it in the excellent book Quantitative Value.

Why might adding in return on capital drag down the strategy's overall performance? It may occur if higher ROCs are likely to be priced in correctly, with the market paying a premium for the highest-quality companies. Also, the best bargains may be eliminated from Magic Formula consideration entirely because of lower ROCs.

The Magic Formula in the Large Cap Space

Interestingly, the Magic Formula performs closer to the raw EV/EBIT strategy when applied to large caps. The following is the performance of the 50 S&P 500 companies with best Magic Formula ranking and benchmark of the equal-weighted S&P 500.

Magic Formula Backtest for S&P 500 Companies

This annualized return of 12.31% is only a moderate decrease relative to the 12.75% EV/EBIT annualized return for large caps.

The following are the 10 highest-ranked S&P 500 companies by the Magic Formula:

  1. Gilead Sciences (NASDAQ:GILD)
  2. Apple (NASDAQ:AAPL)
  3. HP (NYSE:HPQ)
  4. Western Union (NYSE:WU)
  5. Cisco Systems (NASDAQ:CSCO)
  6. Aetna (NYSE:AET)
  7. Michael Kors Holdings (NYSE:KORS)
  8. Humana (NYSE:HUM)
  9. Discovery Communications (NASDAQ:DISCA)
  10. Scripps Networks Interactive (NASDAQ:SNI)

To generate screens and backtests like these, you may sign up for a trial subscription at Portfolio123.


Joel Greenblatt's Magic Formula is a market-beating value strategy. Specifically we have seen that:

  • The Magic Formula has 13.08% annualized returns since 2000 within the Russell 3000.
  • Annualized returns decrease only to 12.31% when applied to a universe of large caps.
  • We can improve on Magic Formula performance by eliminating the return on capital component, and focusing only on earnings yield.

The title of this series is "Best-Performing Value Strategies." The next article, however, will focus on a classic value strategy that appears to have failed remarkably so far this century.

Disclosure: I/we have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.

I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.

About this article:

Author payment: $35 + $0.01/page view. Authors of PRO articles receive a minimum guaranteed payment of $150-500.
Want to share your opinion on this article? Add a comment.
Disagree with this article? .
To report a factual error in this article, click here